The proposal includes three studies designed to clarify measurement of clustering in free recall and apply it to the study of developmental trends in children. A subject-paced procedure is described in which inter-stimulus study times, and inter-stimulus response times will be recorded. These data should directly indicate organizational strategies during acquisition and recall. A critical review of measures of clustering indicates their failure to consider characteristics of the stimulus list. A new measure (X) is presented which computes clustering from recall responses taking into account the categorical characteristics of the stimulus lists. The proposed experiments will study the validity of these measures with children from seven to ten years of age. Each study will include variation of the categorical properties of the stimulus lists within a developmental design and use the subject-paced apparatus. Five dependent variables will be compared. These are: chance clustering, X (new proposed measure), inter-item study time, inter-item response time, total number of items recalled. Study 1 investigates the relationship between the categorization within lists and clustering. Study 2 manipulates category distinctness and compares clustering measures. Study 3 incorporates the findings of the first two studies into a training program to facilitate clustering.